Dose-response relationship among body mass index, abdominal adiposity and atrial fibrillation in patients undergoing cardiac surgery: a meta-analysis of 30 prospective studies

Abstract

AbstractBackgroundWhether being overweight increases the risk of postoperative atrial fibrillation (POAF) is unclear, and whether adiposity independently contributes to POAF has not been comprehensively studied. Thus, we conducted a meta-analysis to clarify the strength and shape of the exposure-effect relationship between adiposity and POAF.MethodsThe PubMed, Cochrane Library, and EMBASE databases were searched for prospective studies (RCTs, cohort studies, and nest-case control studies) reporting data regarding the relationship between adiposity and the risk of POAF.ResultsThirty publications involving 139,302 patients were included. Analysis of categorical variables showed that obesity (RR: 1.39, P<0.001), but not being underweight (RR: 1.44, P=0.13) or being overweight (RR: 1.03, P=0.48), was associated with an increased risk of POAF. In the exposure-effect analysis, the summary RR for a 5-unit increment in body mass index (BMI) was 1.09 (P<0.001) for the risk of POAF. There was a significant linear relationship between BMI and POAF (Pnonlinearity=0.91); the curve was flat and began to rise steeply at a BMI of approximately 30. Notably, BMI levels below 30 (overweight) were not associated with a higher risk of POAF. In the subgroup analysis of surgery types, the pooled RR values for a BMI increase of 5 for coronary artery bypass graft and valve surgery were 1.21 (P<0.01) and 1.34 (P=0.25), respectively, suggesting that a potential difference in the association exists by surgery type. Additionally, waist obesity was associated with the risk of POAF (RR: 1.55, P<0.001).ConclusionBased on the current evidence, our findings show that adiposity was independently associated with an increased risk of POAF, while being underweight or overweight might not significantly increase the POAF risk. The magnitude of the effect of obesity on AF in patients undergoing valve surgery might be small, and this finding needs to be further confirmed.